emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Savannah-hackers] address@hidden: Jan's Gtk CVS notifications...?]


From: Jaime E. Villate
Subject: Re: [Savannah-hackers] address@hidden: Jan's Gtk CVS notifications...?]
Date: Thu, 23 Jan 2003 12:00:46 +0000
User-agent: Mutt/1.2.5i

On Thu, Jan 23, 2003 at 03:00:53AM -0500, Richard Stallman wrote:
> Please take a look at this problem.

> To: address@hidden
> From: Alex Lancaster <address@hidden>
> Date: 21 Jan 2003 18:51:02 -0800
> Subject: Jan's Gtk CVS notifications...?

> It appears that the CVS notification messages for Jan's GTK-related
> CVS check-ins aren't going to the emacs-commit list.
> 
> They're not coming through the GMANE mail-to-NNTP gateway, nor do they
> appear in the archives on mail.gnu.org at:
> 
> http://mail.gnu.org/archive/html/emacs-commit/2003-01/threads.html
> 
> They do appear in CVS (under username "jhd") and the log messages are
> there in the viewcvs page.
> 
> Is there are reason his CVS commit messages aren't going through
> (perhaps you need to be subscribed to the emacs-commit list for the
> commit messages to go to the mailing list)?  To keep up with and test
> the ongoing GTK it is important to get the CVS commit messages from
> all commiters...

Hi Alex,
>From the point of view of Savannah, I can tell you that all commit messages
are sent, regardless of the author's address. I've looked at the CVS
history files and everything looks fine; the problem might be at mail.gnu.org
that receives the e-mail notifications and redistributes them to emacs-commit.
I'm sorry I'm not familiar with the new method used in mail.gnu.org to
distribute mail, so I cannot help you (Paul Fisher <address@hidden> should be
able to help you).

What I can tell you is that there are three Emacs developers (lektu, kfstorm
and handa) whose commits may stop sending e-mail notifications at any time,
because they continue to use the old CVS repository, rather than the new one,
in spite of the warnings given when they commit to the old repository.

The old repository (/home/cvs/emacs or /cvs/emacs) is a symbolic link to the
new one (/cvsroot/emacs), so the files are committed in the same place, but
the CVSROOT files for both repositories are different and they have separate
history files. If the two CVSROOTs ever become out of sync, CVS
may start behaving differently for those 3 developers than for the rest.

Whenever the Emacs maintainers request it, we can remove the link from
/home/cvs/emacs to the new repository, to avoid problems and to keep a unique
history file. What will happen at that time is that users who continue to use
the old one will get an error message telling them the repository does not
exist, but they can find out where the new one is in Savannah's Emacs page.

Cheers,
Jaime





reply via email to

[Prev in Thread] Current Thread [Next in Thread]